Hour Glass Half-Full or Half-Empty? Future Time Perspective and Preoccupation with Negative Events Across the Life Span

Age

The idea that focusing on a limited future may co-occur with focusing on a future full of opportunities is conceptually similar to the gain-loss dynamic posited within life-span developmental theory (Baltes, 1987; Baltes, Lindenberger, & Staudinger, 2006). Life-span theory posits that developmental gains and losses coexist at all ages, but that the relative balance differs by age. For example, research has shown that fluid cognitive abilities start showing age-related decline in early adulthood whereas practical knowledge from life experience continues to improve (Baltes et al., 2006; Li, Baldassi, Johnson, & Weber, 2013; Park, Lautenschlager, Hedden, Davidson, & Smith, 2002; Salthouse, 2011).

People seem to be aware of this age-related shift in the gain-loss dynamic (Ebner, Riediger, & Lindenberger, 2009; Heckhausen & Krueger, 1993). Adults expected gains and losses to be co-occurring with the balance shifting toward losses in older age (Heckhausen, Dixon, & Baltes, 1989; Heckhausen & Krueger, 1993). Younger adults goals' focused more on promoting gains than on preventing losses, whereas older adults' goals focused more on maintaining functioning and preventing losses (Ebner, Freund, & Baltes, 2006). Indeed, young adults may perceive their futures mostly in terms of opportunities due to delaying commitments to careers and relationships until their late 20s or early 30s (see Arnett, 2000). However, it is also possible that some young people perceive life as limited because they don't believe they will live long enough to reap the benefits of their choices (McDade, Chyu, Duncan, Hoyt, Doane, & Adam, 2011; see also Fischhoff, Bruine de Bruin, Parker, Millstein & Halpern-Felsher, 2010 for research with adolescents that supports this idea). Measuring future time perspective as two dimensions could help to clarify how young people balance focusing on limitations and opportunities when thinking about the future.

In midlife, Cate and John (2007) found that increases in perceiving a limited future were not mirrored by decreases in perceived opportunities. Their female participants maintained a focus on future opportunities from the 40s to the 60s, whereas they increasingly focused on a limited future from their 40s to their 50s (Cate & John, 2007). Lachman (2004) theorizes that during midlife (roughly 40-60 years of age), people become increasingly aware that life is limited but simultaneously believe that time remaining is still substantial. Hence, in midlife, focus on future opportunities and limitations may be relatively balanced (see Lachman, Teshale, & Agrigoroaei, 2015).

In later life, there are fewer opportunities to make major changes in one's life path (Neugarten, Moore, & Lowe, 1968). Thus, it is not surprising that older people perceive fewer opportunities than younger people to undo life regrets (Bauer, Wrosch, & Jobin, 2008; Wrosch, Bauer, & Scheier, 2005). Yet, opportunities for volunteering, part-time employment, and leisure increase after ceasing full-time employment (Moen & Flood, 2013). Limitations might not outweigh opportunities until later in life, when it becomes impossible to reverse earlier decisions (Wrosch & Heckhausen, 2002).

Gender

Women may perceive their futures as more limited than men do because, on average, they perceive poorer health and more functional limitations than men (Crimmins, Kim, & Solé -Auró, 2011; Dahlin & Härkönen; 2013; Oksuzyan, Brønnum-Hansen, & Jeune, 2010). Alternatively, men may perceive their future as more limited because their average life expectancy is shorter than women's (e.g., Philips, 2006; Rochelle, Yeung, Bond, & Li, 2015). There is considerable heterogeneity of functioning among both men and women as they age, with many older adults remaining vigorous and active (see Baltes & Smith, 2003; Dannefer, 2003; Neugarten, 1975). As such, age and gender differences in future time perspective may reflect individual differences in health, cognitive functioning, and life events that mark movement through the life cycle.

Retirement

Retirement, in particular, has been suggested to be a developmentally significant life event (e.g., Moen, 2012; Neugarten, 1976). In prior research, expansive time horizons were related to preferring to retire later (van Solinge & Henkens, 2010) and being motivated to achieve while still at work (Kooij & van de Voorde, 2011). Limited time horizons were associated with feeling less obligated to the workplace (Bal, Jansen, van der Velde, de Lange, & Rousseau, 2010) and less achievement motivation (Kooij & van de Voorde, 2011). Research has not addressed whether future time perspective is more limited among those who are retired compared to those who are not, but other life events that signal endings (e.g., college graduation) are related to perceiving a limited future (Ersner-Hershfield, Mikels, Sullivan, & Carstensen, 2008; Fredrickson 1995; Fung, Carstensen, & Lutz, 1999). However, others have argued that a focus on future opportunities may increase after retirement (Parker, Carvalho, & Rohwedder, 2013), perhaps due to having more time for volunteering and leisure (see Moen & Flood, 2013). Measuring future time perspective as two dimensions could help to address how retirement may relate to a seeing the future in terms of both opportunities and limitations.

Perceived health

Older age is associated with worse health (e.g., Ailshire & Crimmins, 2013) and reporting worse subjective health (Liang et al., 2010; Pinquart, 2001). Poorer subjective health relates to perceiving limited future time (Kooij & van de Voorde, 2011), whereas better subjective health relates to having a more expansive time perspective (Bal, Jansen, van der Velde, de Lange, & Rousseau, 2010; Kooij & van de Voorde, 2011; Windsor, Fiori, & Crisp, 2012). Subjective health informs adults' expectations of the likelihood of surviving to age 75 and 85 (Hurd & McGarry, 1995; Hurd, 2009). These expectations predict mortality, perhaps because they capture individuals' knowledge of health changes (Hurd & McGarry, 2002). Perceived changes in health have not been investigated in relation to future time perspective. Perceiving improvements in health could bolster perceptions of future opportunities, whereas perceiving worsening health could intensify a focus on a limited future.

Perceived decision-making ability

Whereas some research has linked future time perspective to subjective health, perceived cognitive functioning has not been investigated. Similar to subjective health (e.g., Kooij & van de Voorde, 2011), perceiving declines in aspects of cognitive abilities with age could increase the salience of a limited future. Decision making is a higher-order cognitive skill that partly depends on basic cognitive abilities (Bruine de Bruin, Parker, & Fischhoff, 2012; Del Missier, Mantyla, & Nillson, 2015; Stanovich & West, 2000; Zaval, Li, Johnson, & Weber, 2015). Older adults rated their decision making as worse than younger adults did, even though not all aspects of decision making declined with age (Bruine de Bruin et al. 2012; see Strough, Bruine de Bruin, & Peters, 2015 for a review). Because theorists have recently suggested that age differences in decision making may be tied to age differences in future time perspective (e.g., Löckenhoff & Rutt, 2015; Mikels, Schuster, & Thai, 2015; Strough, Parker, & Bruine de Bruin, 2015), it is important to establish whether future time perspective is related to people's perceptions of how their decision-making ability has changed as they have gotten older.

Decision-making ability compared to age peers

In addition to judging age-related changes in their decision-making abilities, people may also judge their abilities in comparison to others (Suls & Wheeler, 2012). As people grow older, they perceive their decision-making competence as worse compared to other people whose age is not specified (Bruine de Bruin et al., 2012). However, when comparing to most people their age, they tend to believe they are better off in terms of having fewer life problems and more desirable personality attributes (Bauer, Wrosch, & Jobin, 2008; Heckhausen & Brim, 1997; Heckhausen & Krueger, 1993). Maintaining the perception that one is better off than people the same age can help older adults reduce life regrets (Bauer et al., 2008). Research has not addressed how people perceive their decision abilities compared to same-age peers, or how it relates to future time perspective.

Age

Socioemotional selectivity theory posits that an important consequence of having a limited future time perspective is increased motivation to maximize emotional well-being (Carstensen, 2006; Carstensen et al., 1999). On average, emotional well-being is maintained and even improves across adulthood, declines are not seen until the 70s or 80s (Carstensen et al., 2011; Smith, Borchelt, Maier, & Jopp, 2002; cf., Stanley & Isaacowitz, 2011; for a review, see Charles & Carstensen, 2009). Older adults' ability to maintain emotional well-being despite experiencing relatively more losses than gains has been referred to as the well-being paradox (Baltes & Baltes, 1990).

It has been posited that older adults may maintain emotional well-being through effective emotion regulation (Carstensen et al., 2003; Lawton, 2001; Scheibe & Blanchard-Fields, 2009; Urry & Gross, 2010). Emotion regulation refers to “processes by which individuals influence which emotions they have, when they have them, and how they experience and express these emotions” (Gross, 1998, p. 275; see also Gross & Barrett, 2011 for a discussion of how emotion regulation is distinct from emotion generation). For example, older adults have been found to use preemptive behaviors to avoid negative situations (Charles, 2010; Charles & Luong, 2013; Urry & Gross, 2010).

When they are exposed to negative situations in lab studies, older adults' are often found to be as good, or better, than younger adults at emotion regulation (Larcom & Isaacowitz, 2009; Luong & Charles, 2014; Philips, Henry, Hosie, & Milne, 2008). For example, older adults were better than younger adults at following instructions to focus on the positive aspects of a negative situation (Shiota & Levenson, 2009). However, older adults do not always effectively use specific strategies (Urry & Gross, 2010). Instructions to rethink a negative situation or disengage from it emotionally were less effective for older adults than for young adults (Opitz, Rauch, Terry, & Urry, 2012; Shiota & Levenson, 2009), perhaps because older adults needed more time to execute these strategies (Allard & Kensinger, 2014) or because these strategies taxed fluid cognitive abilities that declined with age (Opitz, Lee, Urry, & Gross, 2014). Older adults who received instructions to remain engaged with a stressor by ruminating about it had delayed physiological recovery, but similar affective recovery, compared to younger adults and an older adult control group (Robinette & Charles, 2014). The latter studies highlight how cognitive and physiological vulnerabilities challenge older adults' capacity for effective emotion regulation (Charles, 2010; Charles & Luong, 2013; Labouvie-Vief, 2003).

Due to these vulnerabilities, one way that older adults can mitigate their exposure to a negative event is to avoid dwelling on the event or becoming preoccupied with it. Older adults report less preoccupation and rumination about negative events, and fewer intrusive and perseverative thoughts (Brose, Schmiedek, Lovden, & Lindenberger, 2011; Kessler & Staudinger, 2009, Nolen-Hoeksema, & Aldao, 2011; Sutterlin, Paap, Babic, Kübler & Vögele, 2012). Dwelling on negative events is linked to prolonged negative affect and physiological stress responses (Ottaviani et al., 2016). Some research suggests that aging dampens the link between intrusive thoughts about stressors and negative affect (Brose et al., 2011) but a recent study that took type of negative affect into account suggested that aging intensified the link—preoccupation with daily stressors increased sadness and disappointment, especially among older adults (Wrzus, Luong, Wagner, & Riediger, 2015). In addition, older adults' lesser preoccupation with negative events explained why they reported less negative affect and more low-arousal positive affect than younger adults (Kessler & Staudinger, 2009).

Whether older adults' lesser preoccupation with negative events reflects their future time perspective is unclear – which may be due to the use of a unidimensional measure in prior research. Kessler and Staudinger (2009) found that people who reported less preoccupation with negative events had more expansive future time perspectives, whereas socioemotional selectivity theory predicts that this type of response, and other responses that increase positive affect, would be associated with having a limited future time perspective, typical of older adults. Ramsey and Gentzler (2014) also found that expansive time horizons were associated with strategies that boosted positive affect. Yeung and colleagues (2012) found that future time perspective explained older adults' lesser endorsement of problem-focused strategies for dealing with interpersonal problems, but did not explain their emotion-regulation strategies. The unidimensional measure of future time perspective used in these studies could have masked a more complex relationship where focusing on limited time also related to emotional reactions to negative events.

If future time perspective consists of two dimensions, it is also possible that each dimension has different associations with emotional reactions depending on age. For instance, Charles and Luong (2013) suggested that the association between focusing on limited time and better affective experience may change during terminal decline—the worsening of physical states and cognition in the three to five years preceding death (see Gerstorf et al., 2010). These dramatic increases in vulnerabilities may limit adaptive capacity to respond to negative events. As noted, cross-sectional and longitudinal studies sometimes find slight declines in positive affect and increases in negative affect during the 70s and 80s (Carstensen et al., 2011; see Charles & Carstensen, 2010 for a review). Hence it is possible that as individuals approach the end of life, focusing on a limited future is associated with less adaptive emotional reactions.

Gender

If, as noted earlier, future time perspective differs depending on gender, then gender could moderate associations between future time perspective and preoccupation with negative events. Women use more strategies to regulate emotion compared to men (Blanchard-Fields, Stein, & Watson, 2004). In a meta-analytic review, women were more likely than men to report using 11 of the 17 strategies investigated, including unproductive rumination and positive reframing (Tamres, Janicki, & Helgeson, 2002; for recent research with similar findings see Garnefski, Teerds, Kraaij, Legerstee & van den Kommer; 2004; Martin & Dahlen, 2005; Nolen-Hoeksema & Aldao, 2011; Zimmermann & Iwanski, 2014; Zlomke & Hahn, 2010). Studies that consider interactions between age and gender are mixed --- some indicate that age differences in strategies depend on gender (Nolen-Hoeksema & Aldoa, 2011), others report similar changes for men and women (Diehl, Chu, Hay, Lumley, Gruhn, & Labouvie-Vief, 2014). These studies highlight gender as an important individual difference characteristic to consider.

Participants

Research Questions 1 and 2 were tested as part of Survey 1. It was completed by 3,933 members of RAND's American Life Panel (ALP), a national sample of adults in the US (https://mmicdata.rand.org/alp/) who regularly receive invitations to complete internet surveys fielded by diverse researchers and earn about $20 per 30 minutes of their time.¹ The invitation to Survey 1 was sent to 4,721 panel members, for a response rate of 83.3%. The sample for Survey 1 consisted of women (59.5%) and men (40.5%) who ranged from 18 to 93 years of age (see Table 1 for demographic information).

Research Question 3 was tested on a subset of 1,045 participants from the larger sample who responded to a second survey that included the measure of reactions to negative events. The second survey was designed for a different study and the invitation went to 1,362 potential participants, for a response rate of 76.2%. The sample for Survey 2 consisted of women (54.8%) and men (43.7%) who ranged in age from 18 to 93 years of age (see Table 1).

Measures

Survey 1 started with questions about participants' experiences as panel members (e.g., number of times they contacted the help desk; types of surveys they liked the most). Measures for this study followed those questions and were presented in the following order: annual health change, future time perspective, perceived changes in decision-making ability with age, and perceived decision-making ability relative to age peers. Self-reported retirement status was from the last quarterly survey the panel member had completed before responding to the measures included in this report.² Survey 2 included the measure of reactions to negative events, after first presenting questions about decision-making style (see Delaney, Strough, Parker, & Bruine de Bruin, 2015). It was in the field at the same time as Survey 1.

1. Is future time perspective better conceptualized as a one- or two-factor construct when investigated in a broad U.S. life-span sample of both women and men?

We first conducted an exploratory factor analysis of the future time perspective items. The scree test indicated only two factors with initial eigenvalues that exceeded 1.0—specifically, 5.50 and 1.97. After PROMAX rotation⁴, the two factors accounted for 45.82% and 16.43% of the variance. An exploratory factor analysis that used only the original 10 items from Carstensen and Lang's (1996) scale also yielded a two-factor solution comprised of focus on future opportunities and limited time. Thus, the two factors were not the result of adding two items to the original scale.

Next, we used a confirmatory factor analysis in Amos to test whether future time perspective was a two-factor construct. To allow a direct comparison with Cate and John (2007), we followed their procedure and did not allow double loadings or correlated errors. The fit of the single-factor model of future time perspective was inadequate, χ²(54) = 117.17, CFI = .75, Δχ² = 3600.25, p < .01, RMSEA = .171, replicating Cate and John's (2007) findings. The two-factor model approached adequate fit (i.e., a CFI cutoff = .90 as recommended by Bentler, 1992, and used by Cate & John, 2007), χ² (53) = 51.45, CFI = .893, RMSEA = .113, but did not meet the more stringent CFI cutoff of .95 advocated by Hu and Bentler (1999). Because model fit of the two-factor model was only adequate, we investigated whether it could be improved by allowing items to covary. We used a more stringent CFI cutoff of .95 (Hu & Bentler, 1999) and RMSEA cutoff of .08 (Byrne, 2010). As shown in Figure 1, allowing items to covary produced a well-fitting model, χ²(47) = 26.19, CFI = .953, RMSEA = .082. Multi-group analysis that compared the fit of the model for men and women confirmed measurement equivalence for the groups.

We labeled the first factor ‘focus on future opportunities.’ Seven of the eight items that loaded on this factor included the word ‘future’ or a word that evoked thoughts of the future (i.e., ‘ahead’), and five referenced either ‘possibilities’, ‘opportunities’, ‘anything’ ‘infinite’ or ‘new goals’. The items that loaded the highest on this factor were ‘my future is filled with possibilities’ followed by ‘many opportunities await me in the future’ and ‘I expect that I will set many new goals in the future’ (see Table 2 for rotated factor loadings of items). We computed a focus on future opportunities score that was the average of eight items that loaded on the first factor. One of these items, “There are only limited possibilities in my future,” was reverse scored. As noted in the methods section, the internal consistency of the scale was acceptable (α = .81 overall, α = .81 for men, α = .81 women; range = 1 – 7, M = 4.62, SD = 1.43). Higher scores indicated greater perceptions of future opportunities.

We labeled the second factor ‘focus on limited time.’ All four items that loaded on this factor included the word ‘time’. Of these four, two items also included the word ‘limited’ and the others included wording that conveyed the idea of having limited time (time ‘running out’, time ‘passing’). The two items that loaded the highest on this factor were ‘As I get older, I begin to experience time as limited,’ and ‘I have the sense that time is running out’ (see Table 2 for rotated factor loadings of items). We computed a focus on limited time score that was the average of the four items that loaded on the second factor. As noted, the scale had acceptable internal consistency (α = .79 overall, α = .79 men, α = .79 women; range = 1 -7, M = 3.97, SD = 1.39). Higher scores indicated greater perceptions of limited time. Scale scores for focus on future opportunities and limited time were significantly correlated (r = -.41, p < .001 overall, r = -.38, p < .001women, r = - .44 p < .001 men).

2. Are there systematic individual differences in future time perspective?

To investigate whether and how age, gender, retirement status, perceived annual change in health, perceived changes in decision-making ability with age, and perceived decision-making ability in comparison to age peers were associated with perceptions of having limited time or future opportunities, we conducted regression analyses.⁵ Multicollinearity indices (tolerance, VIF) were examined and there were no issues found for either multiple regression analysis. Correlations among variables are shown in Table 3, correlations by gender are shown in Table 4. Tests of nonlinear effects of age indicated that for both dimensions of future time perspective, the quadratic function of age was significant and fit the data better than a simple linear function (p < .001).⁶ The quadratic functions indicated that correlations between older age and focusing more on limited time, and less on future opportunities were stronger as age increased (see Figure 2). Due to these nonlinear associations, the age-squared term was included in each regression analysis. Furthermore, up through middle age (until around age 60), people focused more on future opportunities than on limited time, but then the relative focus shifted such that people focused more on limited time than future opportunities.

3. How is future time perspective related to age differences in responses to negative events?

To investigate whether age differences in preoccupation with negative events were accounted for by age differences in future time perspective (focus on future opportunities and limited time), we computed a structural equation model in Amos to test indirect effects. The specified model included a direct association between age and the mean score for avoiding preoccupation, and indirect associations from age through the two latent future time perspective variables to the mean score for avoiding preoccupation. Controlling for correlates of future time perspective (i.e., perceived annual health change, perceived changes in decision-making ability with age, perceived decision-making ability in relation to age peers) did not change the results. The reported coefficients are not adjusted for covariates. This model fit well, X²/df = 3.65, CFI = .950, RMSEA = .050, when judged against criteria outlined by Byrne (2010) and Hu and Bentler (1999). Multi-group analysis that compared the fit of the model for men and women confirmed that model fit was equivalent for the two groups, indicating that all associations between age, focus on limited time, future opportunities, and avoiding preoccupation were the same for men and women.

The constrained structural residuals model was probed to examine the direct and indirect associations. Examination of direct effects showed that, as expected, older age was associated with avoiding preoccupation, greater focus on limited time, and lesser focus on future opportunities (see Figure 3). Focusing more on future opportunities and less on limited time were each associated with avoiding preoccupation. Next, indirect effects were examined. There were significant indirect effects of age to avoiding preoccupation through future opportunities (B = -.05, 95% CI [-.06, -.03], p = .005) and limited time (B = -.02, 95% CI [-.03, -.004], p = .013).¹¹ As shown in Figure 3, after including the two significant indirect paths, the direct effect of older age on avoiding preoccupation (B = .09, p = .01) was even stronger than when not taking these two paths into account (B = .03, p =.005). That is, there was a suppressor effect, where controlling for the paths through the two dimensions of future time perspective amplified the association between age and avoiding preoccupation. Hence, older age was associated with reporting less preoccupation with negative events in spite of older adults' perceptions of fewer future opportunities and more limited time.

Finally, to test whether age moderated the association between future time perspective and preoccupation, we used Hayes's (2013) PROCESS macro with 5,000 resamples. We tested a model where perceptions of limited time was the predictor, and another model where perceptions of future opportunities was the predictor (the dimension of future time perspective not examined as the predictor in a given model was controlled, as was gender). In both models, avoiding preoccupation with negative events was the criterion, and age was a moderator. Age was not a significant moderator in either analysis.¹² Thus, the associations between avoiding preoccupation with negative events and perceiving more future opportunities and less limited time were consistent for people of all ages.

Future time perspective is a two-factor construct when investigated in a broad U.S. life-span sample of both women and men

Findings from our national life-span sample support measuring future time perspective as two dimensions—limited time and future opportunities. Similar to Cate and John (2007), a two-factor model fit the data, whereas a one-factor model did not fit. Our findings build on those of Cate and John (2007), whose oldest participants were middle-aged women, by establishing the fit of a two-factor model in an age-diverse sample comprised of both men and women. Our results show that adults of all ages simultaneously perceive the life-span hour glass in terms of future opportunities and limited time, offering new insights about aging and future time perspective compared to the commonly used unidimensional scale.

Individual differences in future time perspective

Future time perspective is related to age differences in reactions to negative events

Older age was associated with lesser preoccupation with negative events, and this association was the same for men and women. When negative events occur, remaining focused on those events can prolong negative arousal (Ottaviani et al., 2016), which may be especially damaging to older adults due to increased physiological vulnerability to stressors in later life (Charles, 2010; Charles & Luong, 2013; Robinette & Charles, 2014). Our findings, along with those of others (Kessler & Staundinger, 2009; Wrzus et al., 2015), suggest that avoiding preoccupation with negative events after they occur may be a pathway to affective well-being in later life.

Older adults' limited time horizons were related to reporting more preoccupation with negative events. Hence, our findings did not align with socioemotional selectivity theory, which posits that perceptions of limited time motivate older people to maximize emotional well-being in the here and now (Carstensen, 2006). Other researchers have reported results similar to ours when using a unidimensional measure of future time perspective (Kessler & Staudinger, 2009; Ramsey & Gentzler, 2014). A recent study showed that having a limited future time perspective was associated with a maladaptive emotional profile of reporting more negative affect, more depression, and less positive affect—associations opposite to those predicted based on socioemotional selectivity theory (Grühn, Sharifian, & Chu, 2015). In our sample, associations between focusing on limited time and reporting more preoccupation with negative events held across age and gender, suggesting this finding is relatively robust. Other types of emotional reactions and emotion-regulation strategies (Urry & Gross, 2010) may show different results.

Interestingly, in our study the association between older age and less preoccupation with negative events was even stronger after taking into account that older age was associated with greater perceptions of limited time and fewer future opportunities. That is, older adults reported less preoccupation with negative events despite the fact that they saw a more limited future with fewer opportunities. Because prior research has shown that older adults' lesser preoccupation with negative events facilities their emotional well-being (Kessler & Staudinger, 2009; Wrzus et al., 2014) our findings highlight the need to investigate other potential pathways besides future time perspective that help explain why older adults report less preoccupation with negative events. For example, older adults' greater life experience and practice with emotionally-charged situations may help facilitate reactions that boost positive affect and dampen negative affect (Blanchard-Fields, 2007; Scheibe & Blanchard-Fields, 2009).

Limitations and Future Directions

Like all studies, ours has limitations that can be addressed in future research. First, because our study used correlational, cross-sectional data, we were unable to address developmental changes or causality. As a result, the reported age differences and concurrent associations among variables could be cohort specific. The direct and indirect paths we reported must be understood in terms of variance accounted for when statistically controlling (versus not controlling) other variables, not in terms of developmental change and temporal ordering of causal events. Experimental designs could be used in future research to examine whether focusing on future opportunities or limited time changes emotional reactions (see Strough, Schlosnagle, Karns, Lemaster, & Pichayayothin, 2014); a longitudinal-sequential design could be used to disentangle the age-cohort confound.

Second, although we used two items beyond the original ten, the limited time dimension of future time perspective may have lower validity if the smaller number of items doesn't completely capture the construct. The smaller correlation of focus on limited time with age and reactions to negative events compared to focus on future opportunities suggests this could be an issue. Future research could be directed toward further developing the limitations dimension of the scale. In such research, it will be important to examine the divergent validity of the scales because some of the items (e.g., “my future is filled with possibilities”) seem to overlap with other constructs such as optimism and depression, which could obscure understanding of age-related differences. For example, if focus on future opportunities decreases with age, but optimism remains stable, this could help explain the suppression effect we found. Future efforts to develop the scale could also consider creating a domain-specific version to complement the commonly-used domain-general version. For example, time perspective for interpersonal relationships may have different associations with age compared to time perspective for careers, and compared to the domain-general future time perspective we examined.

Third, our findings do not rule out the idea that focusing on limited time facilitates adaptive reactions and strategies through prioritizing emotion-regulation goals (see Carstensen et al., 1999). Goals vary systematically by age and have been linked to strategies in other studies (Berg, Strough, Sansone, Calderone, & Weir, 1998; Luong & Charles, 2014; Strough, Berg, & Sansone, 1996). Research that links future time perspective to emotion-regulation goals, goals to strategies, and strategies to well-being is needed to better understand processes that facilitate emotional well-being in older age (Isaacowitz & Blanchard-Fields, 2012; Scheibe & Carstensen, 2010).

Fourth, we only used one self-report measure of reactions to negative events, the Action Control Scale (Dieffendorf et al., 2000). Although future time perspective did not account for age differences in preoccupation with negative events as measured by this scale, other research has shown it accounts for age differences in strategies for solving interpersonal problems (Yeung et al., 2012). Research using other measures and methods is needed to address whether the dimensions of future time perspective we have uncovered relate to older adults' use of reappraisal, situation selection, and conflict avoidance to promote emotional well-being (see Birdett & Fingerman, 2005; Carstensen et al., 1999; Charles et al., 2009; English & Carstensen, 2014; Shiota & Levenson 2009; Urry & Gross, 2010). In addition, older age is associated with responding in socially desirable ways, which partly explains why older people report better well-being, higher life satisfaction, more positive affect, and less negative affect (Carstensen & Cone, 1983; Soubelet & Salthouse, 2011). This could have contributed to the age differences in reports of reactions to negative events we found when using the Action Control Scale. Controlling for social desirability in future research would help to address this issue.

Conclusions

Our study advances research on aging by showing that that a commonly-used unidimensional measure of future time perspective is better conceptualized as two dimensions. Using this conceptualization, we found that the life-span hourglass was perceived as more full than empty through middle age, but around age 60 this pattern reversed, and limited time loomed larger than future opportunities. Taking this fundamental psychological shift into account may facilitate the design and implementation of effective interventions to promote healthy aging. For example, our findings suggest that interventions that emphasize long-term consequences (e.g., of financial or health behaviors) may be interpreted very differently by someone in their 60s versus 40s, despite increased longevity within the US. Moreover, the unanticipated links between aging, future time perspective, and reactions to negative events uncovered in our research highlight the need for additional empirical and theoretical work directed toward unlocking the well-being paradox.